As described in the previous years, our studies of protein folding have led to the hypothesis that (1) a previously unknown non-covalent interaction exists in the hydrophobic cores of proteins; (2) this new interaction is exothermic and mediated by the contacting groups which form a closed loop (core interaction loop) in the core; and (3) the core loop interaction is sensitive to the detail of the group contact and therefore has the ability to order the core groups. To know more about the nature of this core loop interaction, we have measured KD for the complete set of homologous and hybrid complexes of type 1, 2, 4 and 5 (64 in total number) prepared from horse, tuna, yeast iso-1 and Candida cytochromes c and the mid point of thermal transition of the 695 nm absorption band for selected complexes. The different types represent the different discontinuity sites of the polypeptide chain. The value for KD was found to vary depending on which species of heme or apogragment (or apoprotein) was used or which combination of heme and apogragments (or apoprotein) was used. Analysis of the data and comparison of the amino acid sequences have allowed us to assign 6 mutations responsible for KD changes. Furthermore, the interaction energy affected by such mutations is non-additive. Such non-linear behavior of the interaction is expected for the core loop interaction. Of the 6 mutations affecting KD, 4 were found to be located in the core of the tight channel (see another report). In the previous years we have assigned 3 more folding units to cytochrome c in addition to the right channel. We assume that each of these folding units are associated with a core loop. Furthermore, the present evidence, combined with the studies in the previous years, suggest that the 4 core loops are coalesced in the ground state and segregated at the activated. The present results have allowed us to calculate the core loop coalescence energy separately from the conventional conformational energy. The calculated core loop coalescence energy (-5.25 kcal/mol at 24 degree C) of horse cytochrome c accounts to approximately 72 per cent of folding energy. Furthermore, the calculated perturbation of the core loop coalescence energy (-4.68 kcal/mol at 24 degree C) by Ile 9 to Leu and Leu 64 to Met mutations essentially explains the difference of folding energy between horse and yeast iso-l-cytochrome c found by E. H. Zuniga and B. T. Nall.